COMPENSATORY NEUROFUZZY SYSTEMS WITH FAST LEARNING ALGORITHMS

In this paper, a new adaptive fuzzy reasoning method using compensator y fuzzy operators is proposed to make a fuzzy logic system more adapti ve and more effective. Such a compensatory fuzzy logic system is prove d to be a universal approximator. The compensatory neural fuzzy networ ks built by both control-oriented fuzzy neurons and decision-oriented fuzzy neurons cannot only adaptively adjust fuzzy membership functions but also dynamically optimize the adaptive fuzzy reasoning by using a compensatory learning algorithm. The simulation results of a cart-pol e balancing system and nonlinear system modeling have shown that I) th e compensatory neurofuzzy system can effectively learn commonly used f uzzy IF-THEN rules from either well-defined initial data or ill-define d data; 2) the convergence speed of the compensatory learning algorith m is faster than that of the conventional backpropagation algorithm; a nd 3) the efficiency of the compensatory learning algorithm can be imp roved by choosing an appropriate compensatory degree.